In vivo imaging of autophagy in a mouse stroke model

BASIC research paper

BASIC Research paper

Autophagy 6:8, 1107-1114; November 16, 2010; © 2010 Landes Bioscience

In vivo imaging of autophagy in a mouse stroke model FengFeng Tian, Kentaro Deguchi, Toru Yamashita, Yasuyuki Ohta, Nobutoshi Morimoto, Jingwei Shang, Xuemei Zhang, Ning Liu, Yoshio Ikeda, Tohru Matsuura and Koji Abe* Department of Neurology; Graduate School of Medicine; Dentistry and Pharmaceutical Sciences; Okayama University; Shikata-cho, Okayama Japan

Key words: autophagy, apoptosis, GFP-LC3 Tg mice, in vivo imaging, tMCAO Abbreviations: CBF, cerebral blood flow; DAPI, 4',6-diamidino-2-phenylindole; GFAP, glial fibrillary acidic protein; GFP, green fluorescent protein; LC3, microtubule-associated protein 1 light chain 3; MCA, middle cerebral artery; MCAO, middle cerebral artery occlusion; NGS, normal goat serum; NeuN, neuronal nuclei; PBS, phosphate-buffered saline; PE, phosphatidylethanolamine; Tg, transgenic; tMCAO, transient middle cerebral artery occlusion; TNB, Tris-NaCl-blocking buffer; TUNEL, terminal deoxynucleotidyl transferase-mediated dUTP-digoxigenin nick end labeling

Recent studies have suggested that autophagy is involved in a neural death pathway following cerebral ischemia. In vivo detection of autophagy could be important for evaluating ischemic neural cell damage for human stroke patients. Using novel green fluorescent protein (GFP)-fused microtubule-associated protein 1 light chain 3 (LC3) transgenic (Tg) mice, in vivo imaging of autophagy was performed at 1, 3 and 6 d after 60 min transient middle cerebral artery occlusion (tMCAO). Ex vivo imaging of autophagy, testing of the autophagy inhibitor 3-methyladenine (3-MA), estern blot analysis, immunohistochemistry, terminal deoxynucleotidyl transferase-mediated dUTP-digoxigenin nick end labeling (TUNEL) and fluorescent analyses were performed on brain sections following tMCAO. In vivo fluorescent signals were detected above the ischemic hemisphere through the skull bone at 1, 3 and 6 d after tMCAO, with a peak at 1 d. Similar results were obtained with ex vivo fluorescence imaging. western blot analysis revealed maximum LC3-I and LC3-II expression at 1 d after tMCAO and fluorescence immunohistochemistry demonstrated that GFP-LC3-positive cells were primarily neuronal, not astroglial or microglial, cells. The number of GFP-LC3/TUNEL double-positive cells was greater in the periischemic area than in the core. These results provided evidence of in vivo autophagy detection, with a peak at 1 d, in a live animal model following cerebral ischemia. This novel technique could be valuable for monitoring autophagic processes in vivo in live stroke patients, as well as for clarifying the detailed role of autophagy in the ischemic brain, as well as in other neurological diseases.

Introduction Similar to the ubiquitin-proteasome system, autophagy is considered to play an important role in preventing accumulation of abnormal protein.1 There is increasing interest in being able to examine the relationship between autophagy and stroke. A very recent report showed that stroke could induce activation of autophagy.2 LC3 is a ubiquitin-like protein and is a useful marker of autophagy. LC3 is present in two different forms— one precursor that is cytosolic, LC3-I and another that is phospholipid-conjugated, LC3-II.3 When autophagic stimulus takes place in a cell, LC3-I converts to LC3-II by conjugation of phosphatidylethanolamine (PE). LC3-II is then localized on phagophores, which ultimately develop into autophagosomes.4 The amount of LC3-II highly correlates with the number of autophagosomes3 and increases after hypoxia-ischemia, with a peak at 1 d.5 The protein p62/SQSTM1 is also associated

with autophagosomes, after binding to LC3-II, and is therefore another marker that is utilized to study autophagy.6 GFP-LC3 transgenic mice were created by a transgenic vector containing an enhanced GFP (EGFP)-LC3 cassette inserted between the CAG promoter7 and the SV40 late poly A fragment. In these particular mice, GFP-LC3 is normally expressed in almost all tissues at a baseline level,8 but GFP-LC3 is also further induced under pathological conditions that activate autophagy.9 These mice can also be used for monitoring autophagy in the model of nutrient starvation.8 There are at least three types of cell-death mechanisms, namely, necrosis, apoptosis and transcriptional repressioninduced atypical death (TRIAD),10 with autophagy possibly representing a fourth type. Necrosis is the mechanism where the cell membrane loses its integrity and becomes leaky. Apoptosis is associated with nuclear and chromatin condensation, DNA fragmentation, organelle swelling, cytoplasmic vacuolization

*Correspondence to: Koji Abe; Email: [email protected] Submitted: 04/23/10; Revised: 08/25/10; Accepted: 08/27/10 Previously published online: www.landesbioscience.com/journals/autophagy/article/13427 DOI: 10.4161/auto.6.8.13427 www.landesbioscience.com Autophagy

1107

and nuclear envelope disruption.11 TRIAD is related to general transcriptional repression, induces slowly progressive atypical cell death, and is also involved in neuronal death in Huntington disease and amyotrophic lateral sclerosis (ALS).10 Because apoptosis plays an important role in the neuronal cell death that occurs around the peri-ischemic area, it has attracted attention as the target of neuroprotective therapy. A recent report suggested a possible relationship between autophagy and the apoptotic mechanism of cell death.5 However, there have been few reports that clearly examined such a relationship in the ischemic brain. In this study, therefore, we attempted to detect in vivo fluorescence imaging of autophagy in this particular mice model after transient cerebral ischemia, to identify the fluorescent signal in relation to autophagic markers in brain sections, and to analyze a possible relationship between autophagy and apoptosis. Results Monitoring in vivo autophagy after tMCAO in GFP-LC3 transgenic mice. A fluorescence imaging analyzer detected no fluorescent signal in the hemispheres in vivo on sham-operated head (Fig. 1A, a). At 1 d of reperfusion after the tMCAO, a fluorescent signal was strongly observed over the head of the ischemic hemisphere (Fig. 1A, b). At 3 and 6 d after the tMCAO, the in vivo fluorescence was still detected over the head of the ischemic hemisphere, but the fluorescent signal gradually faded from 1 to 6 d after the reperfusion (Fig. 1A, b–d). Ex vivo autophagy imaging of cerebral slices. Ex vivo autophagy imaging of cerebral slices showed that the fluorescent signal was scarcely visible in 20 μm cerebral slices from the shamoperated brain (Fig. 1A, e and i), whereas the highest fluorescent signal intensities were detected in the lateral cerebral cortex and striatum (Fig. 1A, f, arrows) of the ischemic hemisphere at 1 d after the tMCAO, with gradual decrease from 3 to 6 d after the tMCAO (Fig. 1A, g and h, arrows). These changes were demonstrably shown in a coloring graded image (Fig. 1A, i–l). With microscopy examination, fluorescence was slightly detected in the sham-operated cerebral cortex and in the peri-ischemic area of the ischemic cortex (Fig. 1A, m and q). After the tMCAO, such fluorescent signals became progressively weaker at 1 and 3 d, and finally disappeared at 6 d in the ischemic core of the cerebral cortex (Fig. 1A, n–p). On the other hand, the fluorescent signal was strongly induced at 1 d after the tMCAO in neural cells of the peri-ischemic area (Fig. 1A, r), which then gradually decreased until 3 and 6 d (Fig. 1A, s and t). Double immunofluorescence with GFP-LC3 and autophagosome marker. For double immunofluorescence analysis, anti-LC3 and anti-p62 antibodies were used as the markers of autophagosomes. Figure 1B shows data of colocalization of GFP-LC3 fluorescence and anti-LC3 in neural cells of the periischemic area, both with the peak at 1 d after tMCAO. In detail, both GFP (green) fluorescence, which represents intrinsic signal and LC3 (red) fluorescence, which represents exogenously added antibody, showed many dot signals in the cytoplasm and those merged well. Figure 1C shows data of colocalization of GFP-LC3 fluorescence and anti-p62 in neural cells of the peri-ischemic

1108

area. The peak of the dots of GFP-LC3 was at 1 d after tMCAO. However, the p62 (red) fluorescent signal was reduced at 1 d after tMCAO compared to the sham-treated cells. Western blot analysis. Results of western blot analysis are shown in Figure 2A (a–d) and Figure 3B, b. In Figure 2A, a sham brain, only a single band of LC3-I was detectable with antibody against LC3, but the band of LC3-II was scarcely detectable. Both LC3-I and LC3-II were induced and reached the maximum at 1 d after tMCAO, and then gradually returned to the sham level by 6 d. GFP-LC3 essentially follows the same process (Fig. 2A, a). At 1 d after tMCAO, the band of Beclin 1 reached the maximum (Fig. 2A, c; p < 0.01). On the other hand, only a reduction of band intensity for p62 was detectable at 1 d with antibody against p62 (Fig. 2A, b; p < 0.01). In contrast, the amount of β-tubulin remained at a constant level among sham-treated samples at 1, 3 and 6 d after tMCAO. The band of GFP-LC3-II was significantly blunted by 3 MA (Fig. 2A, d; p < 0.01). As an apoptotic marker, the band of cleaved caspase-3 was strongly induced at 3 d after tMCAO (Fig. 3B, b; p < 0.01). The LC3-I and LC3-II levels reached maximum at 1 d after tMCAO (Fig. 2A, a; *p < 0.05, **p < 0.01). Double immunofluorescence with GFP-LC3 and neural markers. In brain sections at 1 d after tMCAO, both NeuN and GFAP showed many double-positive cells but not in the shamtreated brains (Fig. 2B, upper 12 parts). In contrast, Iba1 immunostaining showed few positive cells 1 d after tMCAO in the ischemic area (Fig. 2B and bottom 3 parts). Number of autophagy-positive cells after tMCAO. The number of double-positive cells with GFP-LC3 (autophagy) and the markers for neuronal, astrocytic or microglial cells in the ischemic core and peri-ischemic area of the affected cerebral hemisphere are shown in Figure 2C. The number of NeuNpositive cells was 143.3 ± 53.0 in the ischemic core and 314.1 ± 41.8 in the peri-ischemic area, that of GFAP were 49.6 ± 23.1 and 143.3 ± 62.8, Iba1 were 38.6 ± 15.1 and 60.6 ± 23.1. The double fluorescence study showed many GFP-LC3positive NeuN cells in comparison with GFP-LC3-positive GFAP cells in the ischemic area (Fig. 2C, **p < 0.05, **p < 0.01, # p < 0.05, ##p < 0.01). Relations of autophagy and apoptosis after tMCAO. Figure 3 shows a double immunofluorescence analysis with GFP-LC3 and markers of apoptosis (TUNEL, cleaved caspase-3). At 1 d after tMCAO, double-positive cells were observed in a small number of neural cells in the ischemic core and in a large number in the peri-ischemic area (Fig. 3A and arrowheads). In quantitative analysis, the numbers of single-positive cells for GFP-LC3 (autophagy) and TUNEL (apoptosis) in the cortex of the ischemic core and peri-ischemic area, respectively, at 1 d after tMCAO in the ischemic core were 122.6 ± 65.8 and 1,458.3 ± 355.4, and in the peri-ischemic area were 453.4 ± 57.0 and 465.7 ± 65.8. Double-positive cells with GFP-LC3 and TUNEL were 61.3 ± 36.4 in the ischemic core and 245.1 ± 53.7 in the peri-ischemic area (Fig. 3B, *p < 0.05, **p < 0.01). The double-positive cells for GFP-LC3 and cleaved caspase-3 were also observed in the ischemic area (Fig. 3B, a, arrowheads).

Autophagy

Volume 6 Issue 8

Figure 1. Monitoring of GFP-fluorescent signal with the observation of (A, a–d) in vivo imaging over the head of the ischemic hemisphere. (A, e–h) Ex vivo imaging in 20-μm cerebral slices from the same brain at 1, 3 and 6 d after tMCAO. (A, i–l) Results of ex vivo imaging were shown as pseudocolor. (A, m–t) Observation of GFP-fluorescence of neuron with microscopy in the core and peri-ischemic area of the ischemic cortex. (B) Double immunofluorescence with GFP-LC3 and autophagosome markers anti-LC3 (red), (C) and anti-p62 (red). Bar 3 μm.

Discussion This is the first report monitoring autophagy in vivo after tMCAO in mice (Fig. 1A). A fluorescent signal was strongly observed over the head of the ischemic hemisphere after tMCAO. This fluorescent signal was detected by the GFP or FITC filter set, but could not be detected using a Cy5 fluorescence filter (data not shown), suggesting that the fluorescent signal actually is related to autophagy. The peak of the in vivo fluorescence signal was at

1 d after tMCAO, and became gradually weaker from 3 d after tMCAO (Fig. 1A, a–d). This result was confirmed by the examination of ex vivo autophagy imaging of tMCAO (Fig. 1A, e–t) showing the same chronological changes. It was reported that the level of GFP-LC3 was comparable to that of endogenous LC3 expression in the brain in GFP-LC3 Tg mice.9 Autophagosomes were labeled with GFP-LC3 during autophagy induction in Tg mice.12 This phenomenon could be observed by fluorescence microscopy.8 Baseline autophagy can

www.landesbioscience.com Autophagy

1109

Figure 2. (A, a–c) Western blot of GFP-LC3, LC3, p62, Beclin 1 and β-tubulin and ratios of LC3/β-tubulin at 1, 3 and 6 d after tMCAO. (A, d) Testing of the 3MA. (B) Double immunofluorescence with GFP-LC3 and neuronal (NeuN), astrocytic (GFAP) and microglial (Iba1) markers. (C) Number of autophagypositive cells after tMCAO. Bar, 3 μm.

be detected in a very small number of GFP-LC3 dots (autophagosomes), but can be further detected in an increasing number of small dots when autophagy is actively induced.8 In our

1110

study, the findings from in vivo imaging suggest that the in vivo fluorescence signal (Fig. 1A, a–d) is related to the formation of the GFP-LC3 dots that are increased in the ischemic cerebral

Autophagy

Volume 6 Issue 8

Figure 3. (A, a) Double immunofluorescence with GFP-LC3 and TUNEL. (A, b) Number of TUNEL and GFP-LC3 double-positive cells at 1 d after tMCAO. (B, a) Double immunofluorescence with GFP-LC3 and cleaved caspase-3. (A, a–c) Western blot of cleaved caspase-3, β-tubulin and ratios of cleaved caspase-3/β-tubulin at 1, 3 and 6 d after tMCAO. Bar, 3 μm.

hemisphere (Fig. 1B and C). In our study, both the intrinsic GFP fluorescence signal and exogenously added antibody of LC3 merged well (Fig. 1B), and in the previous study, it was confirmed that autophagosomes correlate with the GFP signal in the same Tg mice based on electron microscopy.8 This means that the fluorescent signal observed in our study is actually the signal reflecting autophagosome formation. In the previous studies it is shown that 3 MA can inhibit the autophagic process.17 In our study GFP-LC3-II expression in the ischemic area was clearly decreased by treatment with 3 MA compared to treatment with saline at 1 d after tMCAO. In addition, it was reported that p62 is degraded in autophagosomes,13 and therefore it is thought that the level of p62 reflects the autophagic status. In our study the amount of p62 clearly decreased at 1 d after tMCAO (Fig. 2A, b). In addition, immunofluorescence showed the same result. These findings also suggest that autophagy is stimulated after cerebral ischemia. However, the level of p62 was increased at 3 and 6 d after tMCAO, suggesting perhaps that autophagy is upregulated to protect the damaged neurons or that delaying the clearance of autophagosomes causes an increase in components of autophagosomes like p62 (Fig. 1C). LC3 is associated with autophagosome membranes after post-translational modifications. A C-terminal fragment of LC3 is cleaved immediately after synthesis to yield a cytosolic form called

LC3-I. A subpopulation of LC3-I is further converted to an autophagosome-associating form, LC3-II. The amount of LC3-II correlates with the extent of autophagosomes formation.3,14 Thus, the conversion of LC3-I into LC3-II is accepted as a simple method for monitoring autophagy.8,15 Beclin 1 was strongly enhanced in autophagic neurons.16 In our study, we detected the increase of LC3-I, LC3-II and Beclin 1 after tMCAO with western blot analysis with a peak at 1 d (Fig. 2A). These findings suggest the autophagic pathway is activated after tMCAO; concomitantly, expression of LC3-I may increase as a result of upregulation of autophagy. Previous reports show that autophagy can be induced in neurons in the tMCAO model.2 Activation of autophagy increases not only in neurons but also in astrocytes and vascular endothelial cells after tMCAO.2 Our present study showed an activation of autophagy both in neurons and astrocytes, but not in microglial cells (Fig. 2B). The number of autophagy-induced cells in the peri-ischemia area was greater than in the area of the ischemic core, and the number of autophagy-induced neurons was about two times higher compared to the number of autophagic astrocytes at 1 d after tMCAO (Fig. 2C). These results suggest that there are many neurons displaying autophagy in the periischemic area, while many neurons also show apoptosis, implying a relationship between autophagy and apoptosis.

www.landesbioscience.com Autophagy

1111

The level of LC3-II reached the maximum in the area of the ischemic cortex and hippocampus at 6–24 h after cerebral ischemia.5 The number of TUNEL-positive cells reached a peak in the area of the ischemic cortex at 24–48 h after cerebral ischemia.17,18 In our study, although the peak of apoptosis was at 3 d after tMCAO (Fig. 3B, b), autophagy-positive cells were present mainly in the peri-ischemic area and apoptosis-positive cells mainly in the area of the ischemic core at 1 d after tMCAO (Fig. 3B). Furthermore, the number of apoptotic cells was greater than the autophagic cells in the ischemic area at 1 d after tMCAO, suggesting that autophagy could be induced in earlier stages after tMCAO compared to apoptosis. Autophagic and apoptotic mechanisms could occur in the same injured neuron,19 autophagy may inhibit apoptosis,20 or autophagy may promote apoptosis.5 LC3/TUNEL double-positive cells were observed in the ischemic area after tMCAO.2 In our study, the GFP-LC3/TUNEL and GFP-LC3/cleaved caspase-3 double-positive cells were observed in the core and peri-ischemic area after tMCAO (Fig. 3A and B). The number of GFP-LC3 and TUNEL double-positive cells in the peri-ischemic area was about four times higher compared to the ischemic core at 1 d after tMCAO (Fig. 3B). This result suggests that apoptotic and autophagic cell death pathways could occur in the same cells. On the other hand, it was reported that apoptosis would be induced when autophagy was inhibited in neurons,21 and if autophagy is blocked, the cell might quickly proceed towards necrosis by interrupting the apoptotic program which is simultaneously active.16 These result suggests that the autophagic pathway has protective affects against apoptotic cell death and necrosis. However, others report that a sustained high level of autophagy could also lead to cell death.5 In future studies the relationship among autophagy, apoptosis and necrosis should be clarified. In conclusion, this study showed a way of detecting in vivo autophagy in a live animal model. This method could be used as a novel technique for clarifying the role of autophagy not only in the ischemic brain but also in the other neurological diseases. The co-existence of autophagic and apoptotic cell death pathways, found in this study, may be important to define the mechanism of neuronal cell death after tMCAO and to find a better therapy for ischemic stroke patients in the future. Materials and Methods GFP-LC3 transgenic mice. We obtained GFP-LC3#53 mice from the RIKEN Bio-Resource Center in Japan and we maintained this line as homozygotes.9 A transgenic vector of the mice contains an EGFP-LC3 cassette inserted between the CAG promoter (cytomegalovirus immediate-early (CMVie) enhancer and chicken β-actin promoter)7 and the SV40 late polyadenylation signal. In this construction, EGFP is fused to the N terminus of rat LC3B (U05784) so as not to affect C-terminal PE conjugation. The 3.4 kbp CAG-EGFP-LC3-SV40 poly A fragment was microinjected into C57BL/6NCrj BDF1 fertilized oocytes. The initial screen gave 8 transgenic lines. One of them, GFP-LC3#53, was selected because GFP-LC3 was ubiquitously expressed in almost all tissues.8 The GFP-LC3 transgenic neonates could be

1112

distinguished from wild-type neonates using a portable GFP macroscope LEDB-3WOF (Optocode, Japan).9 In the GFP-LC3 Tg mice, GFP-LC3 localizes only on autophagic membranes but not on other membrane structures3 and the GFP-LC3 signal could be detected specifically by the GFP or FITC filter set but not by others.9 When autophagy was induced, the number of GFP-LC3 dots (autophagosomes) observed by microscopy generally increased in vivo.8 The model of transient middle cerebral artery occlusion. The GFP-LC3 8-week-old male mice, weighing 20–23 g, were anesthetized with a nitrous oxide:oxygen:isoflurane mixture (69%:30%:1%) during surgical preparation. The right common carotid artery was isolated and an 8–0 nylon thread with siliconcoated tip was inserted into the right MCA and gently advanced (9.0–10.0 mm). At 60 min after the MCAO, the nylon thread was withdrawn and ischemic brain tissue was reperfused. During the surgery, body temperature was monitored and maintained in the range of 37 ± 0.3°C. Regional cerebral blood flow (CBF) was measured in the ischemic area (1 mm posterior and 3 mm lateral to the bregma) using a laser-Doppler flowmeter (Omegawave, FLO-C1), at the baseline, time point of inducing of ischemia, 5 and 10 min after ischemia and 5 min after the reperfusion, as described in detail previously.22,23 The values of CBF are represented as a percentage of baseline. After the incision was closed, the animals recovered and were allowed free access to water and food at ambient temperature until sampling. The experimental protocol and procedure were approved by the animal committee of the Graduate School of Medicine and Dentistry, Okayama University. Imaging system and in vivo imaging. For the in vivo imaging, we used an Olympus MVX10 in vivo imaging system (Olympus, MVX10) equipped with dual objective (0.63x, 2x), U-RFL-T 100W mercury lamp (Olympus, U-RFL-T) and ORCA-R 2 charge-coupled device camera (Hamamatsu Photonics, C1060010B) and image analysis software MetaMorph (Molecular Devices, Version 7.5). In vivo imaging of autophagy in brain ischemia of GFP-LC3 transgenic mice was performed on day 1, 3 and 6 (n = 7 in 1 d, n = 5 in 3 and 6 d) after 60 min of tMCAO. Animals were anesthetized with a nitrous oxide:oxygen:isoflurane mixture, then the head skin was cut and removed for in vivo detection of GFP light through the parietal bone. The animals were fixed and the body temperature was monitored and maintained in the range of 37 ± 0.3°C inside the Olympus MVX10 system. An in vivo imaging signal was obtained with 0.63x zoom 1. An intensity-controlled laser diode was used for emitting at 488 nm and the U-MGFP was used at 505 nm interference filter (Olympus, U-MGFP/ XL) for the detection of the fluorescence emission and blocking the excitation light. Fluorescence emission was collected by an ORCA-R² digital CCD camera, with exposure time of 300 ms. Image analysis was performed using the above mentioned software MetaMorph. Ex vivo imaging. After obtaining in vivo imaging signals at 1, 3 and 6 d, the same mice were anesthetized with an overdose of pentobarbital (40 mg/kg) and perfused through the heart with 20 ml of ice-cold heparinized saline. The brains were then

Autophagy

Volume 6 Issue 8

removed and rapidly frozen in powdered dry ice. The mice brains were cut on a cryostat into 20-μm thick sections at -18°C and collected on silanized slides (Dako, S3003). The fluorescence of GFP was detected ex vivo in the ischemic area with an Olympus MVX10 using the same condition as for in vivo imaging. Ex vivo images were then analyzed by pseudocolor capture. Immunohistochemistry. The brain sections were washed with phosphate-buffered saline (PBS; pH 7.4) and then blocked for 2 h in 10% normal goat serum (NGS) (Funakoshi, 2v-000210), Tris-NaCl-blocking buffer (TNB) (p62 stain) (Perkin Elmer, 376801) and 0.1% Triton X-100 in PBS. Then they were incubated overnight at 4°C with the primary antibody in 10% NGS, TNB and 0.1% Triton X-100 in PBS. On the next day, the slices were washed in PBS and incubated for 2 h with fluorochrome-coupled secondary antibody (Alexa FluorTM, Molecular Probes, A21424, A21429) at room temperature. The sections were then rinsed three times in PBS and mounted with VECTASHIELD®Mounting Medium with DAPI (Vector Laboratories, H1200). A confocal microscope equipped with argon and HeNe1 lasers (Zeiss, LSM 510) was used to capture fluorescent images. The primary antibodies in this study were as follows: anti-microtubule-associated protein 1 light chain 3 (LC3; 1:400; MBL, PM036), anti-SQSTM1/A170/p62 (1:400; Wako, 018-22141), anti-NeuN (1:400; Chemicon, MAB377), anti-GFAP (1:400; Chemicon, MAB3402), anti-Iba1 (1:400; Wako, 019-19741) and anti-cleaved caspase-3 (1;400; Cell signaling, 9661). Western blot analysis. Western blot analysis was performed using the ischemic hemisphere of three mice from each group. Animals were anesthetized with an overdose of pentobarbital (40 mg/kg) and perfused through the heart with 20 ml of ice-cold heparinized saline. The brains were then removed and the brain tissues were collected from the ischemia area. Then 0.3 mL of cold lysis buffer (50 mM Tris-HCl, pH 7.2, 10% glycerol, 250 mM NaCl, 0.1% NP-40, 2 mM EDTA and protease inhibitors) was added to each tube and it was homogenized at 4°C. The homogenate was centrifuged at 12,000 rpm at 4°C for 10 min and the supernatant fractions (S1) were collected. Protein concentrations of the S1 samples were determined by Lowry assay (Amersham Biosciences, Ultrospec 3100 pro). An amount equivalent to 2 μg of total protein for each sample was subjected to 14% polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membranes (Millipore, IPVH00010). The membrane was References 1.

Morimoto N, Nagai M, Ohta Y, Miyazaki K, Kurata T, Morimoto M, et al. Increased autophagy in transgenic mice with a G93A mutant SOD1 gene. Brain Res 2007; 1167:112-7. 2. Shang JW, Deguchi K, Yamashita T, Ohta Y, Zhang H, Morimoto N, et al. Anti-apoptotic and anti-autophagic effects of GDNF and HGF after transient MCAO in Rats. J Nurs Res 2010; Epub ahead of print. 3. Kabeya Y, Mizushima N, Ueno T, Yamamoto A, Kirisako T, Noda T, et al. LC3, a mammalian homologue of yeast Apg8p, is localized in autophagosome membranes afterprocessing. EMBO J 2000; 19:5720-8. 4. Tanida I, Minematsu-Ikeguchi N, Ueno T, Kominami E. Lysosomal turnover, but not a cellular level, of endogenous LC3 is a marker for autophagy. Autophagy 2005; 1:84-91.

blocked with 5% skimmed milk Tween PBS (skimmed-milk 5 g, 100 ml PBS, pH 7.4, 0.1% Tween-20). We carried out western blot analysis using standard techniques with an ECL plus detection kit (GE Healthcare, RPN 2132) according to our previous report.24,25 The primary antibodies in this study were as follows: anti-LC3 (1:500; MBL, PM036), anti-SQSTM1/A170/p62 (1:200), anti-GFP (1:2,500; MBL, 598), anti-Beclin 1 (1:500; BD, 612113) and anti-cleaved caspase-3 (1:1,000). We carried out densitometry analysis using Scion Image Beta 4.02 software and took the average of the three mice. Cell count. To quantitatively evaluate the results of single immunofluorescence analysis and TUNEL fluorescence, the number of positive cells was counted for NeuN, GFAP, Iba1 and TUNEL in the cortex core and peri-ischemic area of the cerebral cortex, in five coronal sections. In the double fluorescence studies, the double-positive cells were counted in the same manner.26 Terminal deoxynucleotidyltransferase-mediated dUTPdigoxigenin nick end labeling fluorescence. TUNEL assay was performed using an in situ cell death detection kit, TMR red (Roche, 2156792) according to the manufacturer’s instructions. First, the sections were washed with PBS (10 min, three times), they were then incubated with reaction mixture for 60 min at 37°C. The signals were examined using a confocal microscope (Zeiss, LSM 510). The TUNEL fluorescence-positive cells at the ischemic boundary zone were counted using five sections from one brain. Testing of 3-Methyladenine. 3-methyladenine was obtained from Sigma (Sigma, M9281). Intracerebroventricular (icv) injections were given in the ipsilateral ventricle with 2 μl of a 30 mg/ ml solution prepared in saline (0.9% NaCl). Mice were injected at the beginning of reperfusion (n = 4) and sampling was done at 1 d after tMCAO.27 Statistical analysis. Results are shown as the mean ± SD. Obtained data were analyzed by Student’s t-test. A probability value of